Identification of SRY‐box 30 as an age‐related essential gatekeeper for male germ‐cell meiosis and differentiation

Abstract Although important factors governing the meiosis have been reported in the embryonic ovary, meiosis in postnatal testis remains poorly understood. Herein, we first report that SRY‐box 30 (Sox30) is an age‐related and essential regulator of meiosis in the postnatal testis. Sox30‐null mice exhibited uniquely impaired testis, presenting the abnormal arrest of germ‐cell differentiation and irregular Leydig cell proliferation. In aged Sox30‐null mice, the observed testicular impairments were more severe. Furthermore, the germ‐cell arrest occurred at the stage of meiotic zygotene spermatocytes, which is strongly associated with critical regulators of meiosis (such as Cyp26b1, Stra8 and Rec8) and sex differentiation (such as Rspo1, Foxl2, Sox9, Wnt4 and Ctnnb1). Mechanistically, Sox30 can activate Stra8 and Rec8, and inhibit Cyp26b1 and Ctnnb1 by direct binding to their promoters. A different Sox30 domain required for regulating the activity of these gene promoters, providing a “fail‐safe” mechanism for Sox30 to facilitate germ‐cell differentiation. Indeed, retinoic acid levels were reduced owing to increased degradation following the elevation of Cyp26b1 in Sox30‐null testes. Re‐expression of Sox30 in Sox30‐null mice successfully restored germ‐cell meiosis, differentiation and Leydig cell proliferation. Moreover, the restoration of actual fertility appeared to improve over time. Consistently, Rec8 and Stra8 were reactivated, and Cyp26b1 and Ctnnb1 were reinhibited in the restored testes. In summary, Sox30 is necessary, sufficient and age‐associated for germ‐cell meiosis and differentiation in testes by direct regulating critical regulators. This study advances our understanding of the regulation of germ‐cell meiosis and differentiation in the postnatal testis.


| INTRODUC TI ON
Meiosis is a programme of two cell divisions (meiosis I and II), with only one round of DNA replication generating haploid gametes.
Meiosis I is composed of four phases: prophase I, metaphase I, anaphase I and telophase I. Among these phases, prophase I is the longest phase of meiosis and can be divided into five stages: leptotene, zygotene, pachytene, diplotene and diakinesis. The timing and regulation of mammalian meiosis differ dramatically between sexes (Baltus et al., 2006;Feng et al., 2014;Handel & Schimenti, 2010;Spiller & Bowles, 2015). In mouse embryonic ovary, germ cells switch from mitosis to meiosis around 13.5 days post-conception (dpc). In mouse testis, meiosis is initiated around 8-day post-partum (dpp).
Family members of SRY (sex-determining region Y)-box (SOX) are transcription factors containing a highly conserved HMG (high mobility group)-box for DNA-binding (Chew & Gallo., 2009;Gubbay et al., 1990;Sekido & Lovell-Badge, 2009). Over twenty different SOXs have been recognized in mammals. Growing evidence indicates that Sox genes play key roles in diverse embryonic and postnatal developmental processes, including cell fate decision, sex determination and differentiation (Kamachi & Kondoh, 2013;Sarkar & Hochedlinger K, 2013;Sinclair et al., 1990). SRY-box 30 (Sox30), a relatively new SOX member, was first cloned from the mouse testis.
For a long time, Sox30 has been thought to exist only in mammals and its specific role remains unclear. In our previous study, we confirmed for the first time that Sox30 is not only a gene that only exists in mammals, but also exists widely throughout the animal kingdom (Han et al., 2010), suggesting that this gene may be more important than previously considered. Recently, SOX30 was first identified as a novel and key tumour suppressor in lung and ovarian cancers (Han et al., 2015(Han et al., , 2019, and then, as well as in other types of cancers (Guo et al., 2017;Liu et al., 2018;Tao et al., 2019). Our latest study has revealed that epigenetic silencing of SOX30 is a key contributor to male infertility and may offer a therapeutic target for azoospermia in humans (Han et al., 2020). Based on the expression pattern in testis development, Sox30 has been considered likely to be involved in spermatogenesis (Han et al., 2010(Han et al., , 2014Osaki et al., 1999;Petit et al., 2015). Recent research has determined that Sox30 is required for spermiogenesis in mice (Bai et al., 2018;Feng et al., 2017;Zhang et al., 2018). However, it remains unclear whether Sox30 is involved in germ-cell meiosis of postnatal testis.
Our previous study has indicated that Sox30 expression might be associated with germ-cell meiosis in mouse testes (Han et al., 2014).
Herein, we generated an inducible Sox30-null mouse model and first revealed that Sox30-null male mice showed arrested meiosis of spermatocyte and uncontrollable proliferation of Leydig cells, resulting in the complete absence of spermatozoa. At the molecular level, Sox30 dominates male germ-cell meiosis via regulating RA-signalling through direct transcriptional suppression of Cyp26b1, and via direct transcriptional activation of downstream targets Rec8 and Stra8.
Furthermore, Sox30 promotes male differentiation by regulating critical genes of sex differentiation (such as Rspo1, Foxl2, Sox9 and Ctnnb1), especially by direct transcriptional suppression of Ctnnb1.
The defects of germ-cell differentiation and Leydig cell proliferation can be ameliorated in Sox30-null mice after re-expression of Sox30, which is indeed associated with regulation of Cyp26b1, Stra8, Rec8 and Ctnnb1. This study identified Sox30 as a key gatekeeper for male germ-cell meiosis and differentiation, providing new insights into the meiosis of postnatal testis and male differentiation.

| Sox30-Null impairs testes exhibiting abnormal cell differentiation and proliferation
To determine the precise role of Sox30, we generated an inducible Sox30-null mouse model ( Figure S1). Homozygous (Sox30 −/− ) and heterozygous (Sox30 −/+ ) mice developed normally and lived more than one year without an obvious difference in general physical appearance when compared with their wild-type (Sox30 +/+ ) littermates.

| The impairments of testes become more severe in Sox30-null aged mice
Smaller testes were observed in Sox30 −/− male mice when compared with Sox30 +/+ and Sox30 −/+ mice. Interestingly, the loss of Sox30 −/− testicular weight was more severe as mice aged from 1.5 months to 8 months post-partum ( Figure 2a). Pathological examination showed that complete absence of spermatozoa and elongated spermatids was observed in testes of Sox30 −/− mice at all nine developmental stages (1.5, 2, 3, 4, 5, 6, 8, 10 and 12 months post-partum), F I G U R E 1 Sox30-null specifically impairs male fertility and testis with abnormal cell differentiation and proliferation. (a) Fertility was analysed in different Sox30 genotypes of mice. "+/+" represents wild-type mice (Sox30 +/+ ), "-/+" represents heterozygous mice (Sox30 −/+ ) and "-/-" represents homozygous mice (Sox30 −/− ). (b) Testicular morphology was evaluated in Sox30 +/+ , Sox30 −/+ and Sox30 −/− adult mice. "***" represents a p-value less than 0.001. The scale bar is 0.5 cm. (c) The testis weight and testis weight/body weight were analysed in Sox30 +/+ , Sox30 −/+ and Sox30 −/− adult mice. "***" represents a p-value less than 0.001. (d) Haemotoxylin-eosin staining of sections of epididymide and testis are shown in Sox30 +/+ (n = 8), Sox30 −/+ (n = 8) and Sox30 −/− (n = 8) mice at 4 months post-partum. No spermatozoa in Sox30 −/− epididymides and complete absence of spermatozoa with few spermatids and some spermatocyte-like cells in seminiferous tubules of Sox30 −/− testes can be observed. The arrows represent spermatozoa in epididymides. The arrows with "Sp" represent the spermatogonia, the arrows with "So" represent the spermatocytes, the arrows with "Sc" represent the Sertoli cells, the arrows with "R" represent the round spermatids, the arrows with "El" represent the elongated spermatids, and the arrows with "Mn" represent multi-nucleated cells in testes. Scale bars are 50 µm. (e) Leydig cells in testes were compared between Sox30 +/+ (n = 8), Sox30 −/+ (n = 8) and Sox30 −/− (n = 8) adult (4 months) mice. The cell numbers were quantified based on an average of 5 random fields, and the results were normalized by the area and number of seminiferous tubules in the area. The arrows represent Leydig cells. "***" represents a p-value less than 0.001. Scale bars are 50 µm and round spermatids were observed in testes of Sox30 −/− mice at and before 6 months post-partum (Figure 2b,c). Evidently, round spermatids in testes of Sox30 −/− mice decreased in continuity from 1.5 months to 6 months post-partum, and completely disappeared in testes of Sox30 −/− mice at and after 8 months post-partum (Figure 2b,c). The above results were further confirmed in independent samples by semithin histological section and expression analysis of spermatid markers: Tnp1, Tnp2 and Prm2 (Tnp1 and Tnp2 are spermatid markers, and Prm2 is an elongated spermatid marker) (Figure 2d,e). These data revealed that impairments of Sox30-null on testes become more severe as mice aged, suggesting a major impairment in prophase of germ-cell differentiation occurred at least before round spermatid formation.

| The germ cells are arrested at early meiotic prophase in Sox30-null testes
On analysing transcriptomes of testes from Sox30 +/+ , Sox30 −/+ and To further confirm the above hypothesis and determine the arrest phase, we then examined Sox30 expression in postnatal testes from 6 to 12 dpp (a key time period of meiosis) old wild-type C57BL/6 mice. During this period, Sox30 expression started increasing in testes of mice at 8 dpp and peaked at 10 dpp, which is a critical time point for spermatocyte meiosis (Figure 3e). In wild-type C57BL/6 mice, the first spermatogenic meiosis occurs at 8 dpp, and then progresses to leptotene, zygotene and pachytene stages at 10, 12 and 14 dpp, respectively. Moreover, Sox30 expression is mainly restricted to spermatocytes (Figure 3f). These results revealed that the arrest of germ-cell meiosis is associated with leptotene or zygotene stage spermatocytes in Sox30 −/− mice.

| The germ-cell meiosis is arrested at the zygotene stage in testes of Sox30-null mice
To determine the precise phase of meiotic spermatocyte arrest in  Figure S4A,B). Consistent with the arrest of meiosis at zygotene (start at 12 dpp), the difference in testis weight presented no statistical significance among Sox30 +/+ , Sox30 −/+ and Sox30 −/− mice at 8, 10 and 12 dpp, whereas testis weight was lower in Sox30 −/− mice than in Sox30 +/+ and Sox30 +/− mice at 14 and 16 dpp (p < 0.05, Figure S4C). In zygotene, the key event is that homologous chromosomes start to pair and synapse, initiating the assembly of the synaptonemal complex (Lee & Hirano, 2011). In mice, homologous chromosomes start to synapse using a DNA doublestrand break (DSB)-dependent mechanism (Cahoon & Hawley, 2016). Thus, to further confirm the arrest of meiosis at zygotene in Sox30 −/− mice, we determined the expression of DSB marker γ-H2A.X, and found that it was poorly detectable in the testes of Sycp3 (a marker of primary spermatocyte) confirmed that spermatocytes were indeed increased in testes of Sox30 −/− mice ( Figure 4D).
These data demonstrated that the zygotene-to-pachytene transition of germ-cell meiosis was defective in Sox30 −/− testes.

| The meiotic defect of Sox30-null mice is associated with critical regulators of meiosis and sex differentiation
To address the mechanism underlying the arrest of meiosis, we reanalysed the transcriptomes of testes from Sox30 +/+ , Sox30 −/+ and  (8 M), 10 months (10 M) and 12 months (12 M) post-partum. Round spermatids are still observed in Sox30 −/− testes at and before 6 months post-partum. The complete absence of round spermatids and a significant reduction in spermatocytes can be observed in Sox30 −/− testes after 6 months post-partum. The arrows with long tails represent multi-nucleated cells, and the arrows with short tails represent round spermatids. The scale bars are 50 µm. (d) Semithin histological sections stained with toluidine blue were performed in testes of Sox30 +/+ , Sox30 −/+ and Sox30 −/− mice at 6 months (6 M) and 8 months (8 M) post-partum. The arrows with "Sp" represent the spermatogonia, the arrows with "So" represent the spermatocytes, the arrows with "Sc" represent the Sertoli cells, the arrows with "R" represent the round spermatids and the arrows with "El" represent the elongated spermatids. Scale bars are 50 µm. (E) Heatmap of relative expression was analysed for the markers of round spermatids and elongated spermatids in testes of Sox30 +/+ , Sox30 −/+ and Sox30 −/− mice at 4 months (4 M), 6 months (6 M) and 8 months (8 M) post-partum by RT-qPCR level as Sox30 acts as a transcription factor. Bioinformatic analyses revealed that Cyp26b1, Rec8, Stra8, Sox9, Foxl2 and Ctnnb1 promoters contain Sox30 binding sites (Table S1). The full-length promot- Haematoxylin-eosin staining was performed in the testes of Sox30 +/+ , Sox30 −/+ and Sox30 −/− mice. Some multi-nucleated spermatogenic cells are present in the testes of Sox30 −/− adult (2 months) mice. The arrows with long tails indicate spermatocytes with condensed chromatin, and the arrows with short tails represent multi-nucleated cells. Scale bars are 50 µm. (c) Spermatogenic cell proliferation and apoptosis were measured by 5-ethynyl-2'-deoxyuridine (EdU) and TUNEL assays in adult (2 months) testes. The cell numbers were quantified based on an average of 5 random fields and was normalized by the area and number of seminiferous tubules in the area. The mice were intraperitoneally injected with 100 µg of EdU in phosphate-buffered saline (PBS), and spermatogenic cell proliferation assays were performed after incorporation for 72 h. TUNEL positive apoptotic cells are stained brown. The "NS (no significance)" represents a p-value more than 0.05. Scale bars represent 50 µm. (d) The DNA content of testicular cells was determined by flow cytometric analyses in testes of Sox30 +/+ , Sox30 −/+ and Sox30 −/− adult (4 months) mice. The 1C (one copy of chromosome) peak was determined by comparison with sperms of the epididymis. "**" represents a p-value less than 0.01. (e) The expression of Sox30 was evaluated in testes of 6 days post-partum (dpp), 8 dpp, 10 dpp and 12 dpp old wild-type mice. Gapdh was used as an internal control. "**" represents a p-value less than 0.01. (f) Sox30 expression was analysed in the testes of wild-type adult (4 months) mice by immunofluorescence staining. Sox30 is mainly expressed in spermatocytes and marginally expressed in round spermatids. As a negative control, a primary antibody was replaced with PBS. Scale bars are 20 µm The expression of DSB marker γ-H2AX and Sycp3 was determined by western blotting or immunofluorescence staining in testes of Sox30 +/+ , Sox30 −/+ and Sox30 −/− adult (3 months) mice. Relative expression was quantified based on mean grey value of an average of 5 random fields. "**" represents a p-value less than 0.01. "NS" represents a p-value more than 0.05 C-terminus region was deleted ( Figure 6D; Figure S6C). Sox30 activated Stra8 promoter activity when the HMG-box domain and C-terminus region were both deleted or when only the C-terminus region was deleted ( Figure 6D; Figure S6C). These data suggested that the C terminus of Sox30 is required for attenuating Cyp26b1 and Ctnnb1 promoter activity, and the HMG-box of Sox30 is required for stimulating Rec8 promoter activity, while the amino-terminal (N-terminus) region (1-365AA) of Sox30 may be required for activating Stra8 promoter activity, as neither the HMG-box nor the C terminus of Sox30 was required.

| Sox30 is necessary and sufficient for germcell meiosis and differentiation that associated with Cyp26b1, Stra8, Rec8, and Ctnnb1
To determine the recovery of germ-cell meiosis and differentiation after re-expression of Sox30 in male Sox30 −/− mice, we restored Sox30 expression in adult Sox30 −/− mice (Sox30 loxp/loxp ) by deleting F I G U R E 6 Sox30 regulates Stra8, Rec8, Cyp26b1 and Ctnnb1 by direct binding to their promoters. (a) The inhibition and activation of Cyp26b1, Rec8, Stra8, Sox9, Wnt4, Ctnnb1 and Rspo1 promoters were determined using luciferase reporter assays in NIH3 T3 cells. The results were normalized with internal controls and are presented as average with the standard error of the mean (SEM) from three experiments. C26b1-pm represents Cyp26b1 promoter, Rec8-pm represents Rec8 promoter, Stra8-pm represents Stra8 promoter, Sox9-pm represents Sox9 promoter, Wnt4-pm represents Wnt4 promoter, Ctnnb1-pm represents Ctnnb1 promoter and Rspo1-pm represents Rspo1 promoter. "**" represents a p-value less than 0.01. (b) Chromatin immunoprecipitation analyses of Sox30 binding to the promoters of Cyp26b1, Rec8, Stra8 and Ctnnb1 were performed in testes of Sox30 +/+ , Sox30 −/+ and Sox30 −/− adult (3 months) mice or transfected NIH3 T3 cells. The immunoprecipitates were analysed by RT-PCR. C26b1-pm represents Cyp26b1 promoter, Rec8-pm represents Rec8 promoter, Stra8-pm represents Stra8 promoter and Ctnnb1-pm represents Ctnnb1 promoter. The "Vec" represents an empty vector. "M" represents DNA marker. (c) The vectors without HMG-box and C-terminal domain or without only C-terminal domain of Sox30 were constructed. Sox30-Dele-1 and Sox30-Dele-2 represent the plasmid without HMG-box and C-terminal regions, and the plasmid without C-terminal region, respectively. (d) The effect of Sox30 with or without HMG-box and/or C-terminal domains on Cyp26b1, Rec8, Stra8 and Ctnnb1 promoter activities was analysed. C26b1-pm represents Cyp26b1 promoter, Rec8-pm represents Rec8 promoter, Stra8-pm represents Stra8 promoter and Ctnnb1-pm represents Ctnnb1 promoter. "**" represents a p-value less than 0.01 F I G U R E 7 Sox30 is required for testis development and germ-cell meiosis via regulation of Cyp26b1, Stra8, Rec8, and Ctnnb1. (a) The morphology and testis/body weight were analysed in Sox30 +/+ (n = 5), Sox30 −/− (n = 5), and Sox30 loxp/loxp (n = 5) mice. +/+, Sox30 +/+ mice injected with Tam; -/-, Sox30 −/− mice injected with solvent; loxp/loxp, Sox30 −/− mice injected with Tam. "***" represents a p-value less than 0.001. (b) H&E staining of the testes and epididymides was performed in Sox30 +/+ , Sox30 −/− and Sox30 re-expressed Sox30 −/− (Sox30 loxp/loxp ) mice (the recovery time is 4 months). Numerous spermatids and spermatozoa are observed in the testes and several many spermatozoa are detected in the epididymides of Sox30 +/+ mice (n = 5) after Tam injection, but are not observed in the testes and epididymides of Sox30 −/− (n = 5) mice after solvent injection. A considerable amount of spermatids and spermatozoa are detected in the testes of Sox30 loxp/loxp mice, and some spermatozoa can be found in the epididymides of these Sox30 loxp/loxp mice (n = 5). The arrows represent spermatozoa in epididymides. The arrows with "R" represent the round spermatids, and the arrows with "El" represent the elongated spermatids in testes. Scale bars are 100 µm (c). Leydig cells were quantified and compared in the testes of Sox30 +/+ , Sox30 −/− and Sox30 loxp/loxp mice. The cell numbers were quantified based on an average of 5 random fields, and the results were normalized by the area and number of seminiferous tubules in the area. The arrows represent Leydig cells. "***" represents a p-value less than 0.001. "**" represents a p-value less than 0.01. Scale bars are 50 µm. (d) Fertility was analysed in Sox30 −/− mice after re-expression of Sox30 at 2, 4 and 8 months. +/+, Sox30 +/+ mice injected with Tam; -/+, Sox30 −/+ mice injected with Tam; -/-, Sox30 −/− mice injected with solvent; loxp/loxp, Sox30 −/− mice injected with Tam. (e) Expression levels of Sox30, Cyp26b1, Stra8, Rec8 and Ctnnb1 were analysed by western blotting in the testes of Sox30 +/+ , Sox30 −/− and Sox30 loxp/loxp mice. β-actin was used as an internal control. (f) A schematic illustration of Sox30 regulation of germ-cell meiosis is shown. In postnatal testes, following Sox30 expression, Cyp26b1 expression is inhibited, and the retinoic acid (RA) level is increased, with decreased catabolism by Cyp26b1 promoting Stra8 and Rec8 expression or Sox30 directly promotes Stra8 and Rec8 expression; meiosis is in progress and spermatogenesis is normal. On silencing Sox30 expression, Cyp26b1 is highly expressed without inhibition of Sox30, the RA level is decreased, with increased catabolism by Cyp26b1 leading to downregulation of Stra8 and Rec8 expression or Sox30-null directly results in downregulation of Stra8 and Rec8 expression; meiosis is arrested, and spermatogenesis is abnormal the floxed cassette using tamoxifen (Tam), which can activate the ER-Cre and evaluated the changes in testis size, germ-cell differentiation, Leydig cell proliferation and male fertility. The testis size and weight of Sox30 loxp/loxp mice were restored when compared with those of Sox30 −/− mice injected with the solvent ( Figure 7A; Figure S8A).
These data revealed that Sox30 re-expression in adult Sox30 −/− mice can successfully reverse the pathological damage in testes, and restore germ-cell meiosis, differentiation and Leydig cell proliferation.
To analyse whether the functional recovery of testis phenotypes and germ-cell differentiation is associated with regulation of Cyp26b1, Stra8, Rec8 and Ctnnb1 in Sox30 loxp/loxp mice, we evaluated the expression of these proteins in testes from the inducible system.
Rec8 and Stra8 were reactivated, and Cyp26b1 and Ctnnb1 were reinhibited on restoring Sox30 expression ( Figure 7E; Figure S8B). These results suggested that Sox30 is required for germ-cell meiosis and differentiation by regulation of Cyp26b1, Stra8, Rec8 and Ctnnb1.

| DISCUSS ION
Meiosis is a programme with two cell divisions but only one round of DNA replication to generate haploid gametes. In mammals, the timing and regulation of meiosis differ between the two sexes (Baltus et al., 2006;Feng et al., 2014;Handel & Schimenti, 2010;Spiller & Bowles, 2015). In the mouse embryonic ovary, the switch from mitosis to the meiosis of germ cells occurs at approximately 13.5 dpc. In mouse testis, the meiosis is initiated at approximately 8 dpp. Despite the recent identification of some meiosis inducing factors such as RA and Msx1/2 (Bowles et al., 2006;Koubova et al., 2006;Le Bouffant et al., 2011;Raverdeau et al., 2012), it remains largely unclear how meiosis is specifically promoted and maintained in postnatal testes (Feng et al., 2014;Handel & Schimenti, 2010;Rossitto et al., 2015).
In the present study, we identified Sox30 as a male-specific intrinsic factor that promotes or maintains germ-cell meiosis, which partially explains the long-standing issue of how meiosis is specifically promoted or maintained in postnatal testes.
In Sox30 −/− mice, no spermatozoa and no elongated spermatids were found in testes at different developmental stages, whereas round spermatids were observed in testes at and before 6 months.
These phenotypes appeared consistent with the results of previous studies (Bai et al., 2018;Feng et al., 2017;Zhang et al., 2018).
However, in the present study, we observed a considerably different issue, in which round spermatids in Sox30 −/− testes continuously decreased as mice aged from 1.5 months to 6 months post-partum and completely disappeared at and after 8 months post-partum.

This different result indicates that Sox30 acts primary roles in the
prophase of germ-cell differentiation, which probably occurs at least before round spermatid formation. Furthermore, we revealed that Sox30 deletion in mice results in spermatogenic cell arrest at the meiotic prophase; this arrested phase is the zygotene stage of spermatocytes. In zygotene, homologous chromosomes initiate the assembly of the synaptonemal complex (Lee & Hirano, 2011). Thus, it is reasonable to speculate that Sox30 might disrupt the synaptonemal complex during meiosis of male germ cells. However, the exact role of Sox30 in the synaptonemal complex remains unknown, and further studies are required to explore this issue.
Recent investigations have revealed that Sox30 is essential for spermatogenesis (Bai et al., 2018;Feng et al., 2017;Zhang et al., 2018). However, the position of Sox30 expression in testes and the arrest period of germ-cell development in Sox30-deficient mice are controversial. Zhang et al. have shown that Sox30 is predominantly expressed in spermatids , whereas Bai et al. have observed that Sox30 expression is restricted to meiotic spermatocytes and postmeiotic haploids (Bai et al., 2018). Feng et al. and Zhang et al. have revealed that germ cells are arrested at the postmeiotic round spermatids in Sox30 knockout mice (Feng et al., 2017;Zhang et al., 2018), whereas Bai et al. have indicated the accumulation of meiotic diplotene spermatocytes, suggesting an impaired transition from meiosis to postmeiosis in Sox30 knockout mice (Bai et al., 2018). In our present study, we observed that Sox30 expression is predominantly restricted to meiotic spermatocytes (high level) and round spermatids (relatively low level), especially to meiotic spermatocytes, using two Sox30 antibodies, which is additionally confirmed by another recent study (Roumaud et al., 2018).
Particularly, we revealed that Sox30 is a key factor for promoting or maintaining male germ-cell meiosis, and deletion of Sox30 resulted in impaired zygotene-to-pachytene transition. Interestingly, deletion of Sox30 does not cause an immediate arrest of meiosis completely but becomes more apparent and severe in ageing mice. The specific reasons for this phenomenon need to be further investigated.
In the present study, the conclusion of the meiotic arrest of germ cells is distinct from previously reported in Sox30 −/− mice (Bai et al., 2018;Feng et al., 2017;Zhang et al., 2018), but the arrest at zygotene of male germ cells in Sox30 −/− mice is highly credible: (1) the number of round spermatids was not increased as mice aged in Sox30-null testes, and instead, it was decreased and even completely disappeared in ageing mice. If the germ-cell arrest occurs at round spermatids in Sox30 −/− testes, the round spermatids in Sox30 −/− testes should be accumulated at least not decreased or even disappeared; (2) DNA content analyses showed few cells harbouring one copy of the chromosome, and round spermatids also had one copy of chromosome; (3)  Our previous results have indicated that silencing SOX30 is possibly a crucial contributor to azoospermia disease (Han et al., 2020).
However, the precise pathogenesis and mechanism of azoospermia induced by SOX30 silencing remain unclear. Based on the findings of the current investigation, the pathogenesis of azoospermia caused by SOX30 silencing could be that SOX30-loss causes the abnormal arrest of germ-cell meiosis at zygotene, leading to complete absence of spermatozoa and male infertility by directly regulating critical regulators of meiosis (Cyp26b1, Stra8 and Rec8) and sex differentiation (Sox9, Foxl2, Wnt4, Ctnnb1 and Rspo1). Thus, our data uncovered the pathogenesis of human azoospermia and advanced the current understanding of this disorder.
In the present study, the most interesting data generated involved experiments reversing the phenotype using an inducible knockout model. The restoration of germ-cell differentiation and reappearance of spermatozoa were observed following inducible Sox30 expression in adult Sox30 −/− mice. However, it appears that despite analysis at 8 months post-Tam (several complete cycles of spermatogenesis), a marked difference in testis phenotype persists between Sox30 +/+ and Sox30 loxp/loxp mice. This could be mainly attributed to the incomplete restoration of Sox30 in Sox30 loxp/loxp mice owing to the efficiency of recombination. Further expression analyses revealed that the Sox30 expression was incompletely restored in testes of Sox30 loxp/loxp mice. These data suggest that testis phenotypes could be further ameliorated if the efficiency of recombination is improved in the restoration experiments.
In addition, our study revealed that the testis weight was significantly lower in Sox30 −/− mice than in Sox30 +/+ mice, starting from 14 dpp, and Sox30 can regulate several critical genes of sex determination and differentiation, including the downregulation of Sox9 and Sry, and upregulation of Foxl2, Wnt4, Ctnnb1 and Rspo1 from the RNA sequencing data in testes of Sox30 +/− mice. Further experiments revealed that Sox30 significantly enhanced the activity of Sox9 promoter and attenuated the activities of Wnt4, Rspo1 and Ctnnb1 promoters. Moreover, Sox30 suppressed Ctnnb1 expression by direct binding to its promoter, and Ctnnb1 was reinhibited on restoring Sox30 expression in Sox30 −/− mice. These data indicate that Sox30 is strongly associated with sex differentiation even sex determination in mice. Additional investigations are required to clarify the sex ratio of the mutant mice, and the precise role and mechanism of Sox30 on sex differentiation even sex determination at early developmental stages.
In summary, our study revealed that Sox30 plays an age-related key role in promoting or maintaining the meiosis of germ cells in testes, and in sex differentiation by directly regulating Stra8, Rec8, Cyp26b1 and Ctnnb1, and re-expression of Sox30 in adult Sox30 −/− mice can successfully restore germ-cell meiosis and differentiation, and testis development ( Figure 7F). This study advances our understanding of the regulation of male germ-cell meiosis and differentiation, and the pathogenesis of azoospermia.

| Cell lines
The cell lines were obtained from the Cell Bank of the Chinese Academy of Science (CBCAS, Shanghai, China). Details are provided in Supplemental experimental procedures.

| Fertility assays
Breeding assays were carried out, and details are provided in Supplemental experimental procedures.

| Chromosome spread and immunostaining
Chromosome spread assays were performed as previously described (Peters et al., 1997). Detailed methods are provided in Supplemental experimental procedures.

| Flow cytometry
Testicular cell suspensions were prepared as previously described (Rodríguez-Casuriaga et al., 2009. Detailed methods are provided in Supplemental experimental procedures.

| Terminal deoxynucleotidyl transferase-mediated dutp nick end labelling assay
Testicular paraffin sections were prepared and the apoptotic cells were detected using in situ cell death detection kit, POD (Roche, Penzberg, Germany). Details are provided in Supplemental experimental procedures.

| Spermatogenic cell proliferation assay
Spermatogenic cell proliferation was measured by performing the 5-ethynyl-2'-deoxyuridine (EdU) assay. Detailed methods are provided in Supplemental experimental procedures.

| RNA extraction, library construction and transcriptome sequencing
Detailed methods are provided in Supplemental experimental procedures.

| Heatmap analysis
Detailed methods are provided in Supplemental experimental procedures.

| Western blotting
Western blotting (WB) analysis was performed as previously described (Han et al., 2015). Detailed methods are provided in Supplemental experimental procedures.

| Immunofluorescence
Detailed methods are provided in Supplemental experimental procedures.

| Construction of vectors and cell transfection
Detailed methods are provided in Supplemental experimental procedures.

| Site-Directed mutagenesis assay
Sox30 HMG-box constructs were mutated using a QuikChange Lightning Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA, USA) as previously described . Detailed methods are provided in Supplemental experimental procedures.

| Luciferase reporter assay
Detailed methods are provided in Supplemental experimental procedures.